Self-supervised learning has gained popularity because of its ability to avoid the cost of annotating large-scale datasets. It is capable of adopting self-defined pseudolabels as supervision and use the learned representations for several downstream tasks. Specifically, contrastive learning has recently become a dominant component in self-supervised learning for computer vision, natural language processing (NLP), and other domains. It aims at embedding augmented versions of the same sample close to each other while trying to push away embeddings from different samples. This paper provides an extensive review of self-supervised methods that follow the contrastive approach. The work explains commonly used pretext tasks in a contrastive learning setup, followed by different architectures that have been proposed so far. Next, we present a performance comparison of different methods for multiple downstream tasks such as image classification, object detection, and action recognition. Finally, we conclude with the limitations of the current methods and the need for further techniques and future directions to make meaningful progress.

We use a low-dimensional linear model to describe the user rating matrix in a recommendation system. A non-negativity constraint is enforced in the linear model to ensure that each user's rating profile can be represented as an additive linear combination of canonical coordinates. In order to learn such a constrained linear model from an incomplete rating matrix, we introduce two variations on Non-negative Matrix Factorization (NMF): one based on the Expectation-Maximization (EM) procedure and the other a Weighted Nonnegative Matrix Factorization (WNMF). Based on our experiments, the EM procedure converges well empirically and is less susceptible to the initial starting conditions than WNMF, but the latter is much more computationally efficient. Taking into account the advantages of both algorithms, a hybrid approach is presented and shown to be effective in real data sets. Overall, the NMF-based algorithms obtain the best prediction performance compared with other popular collaborative filtering algorithms in our experiments; the resulting linear models also contain useful patterns and features corresponding to user communities.

Supplementary material is available from http://www.cs.dartmouth.edu/~wyh/hykgene/supplement/index.htm.

No abstract

Abstract. The analysis of brain activations using functional magnetic resonance imaging (fMRI) is an active area of neuropsychological research. Standard techniques for analysis have traditionally focused on finding the most significant areas of brain activation, and have only recently begun to explore the importance of their spatial characteristics. We compare fMRI contrast images and significance maps to training sets of similar maps using the spatial distribution of activation values. We demonstrate that a Fisher linear discriminant (FLD) classifier for either type of map can differentiate patients from controls accurately for Alzheimer's disease, schizophrenia, and mild traumatic brain injury (MTBI).

No abstract

Sensor networks are used in a variety of application areas for diverse problems from habitat monitoring to military tracking. Whenever they are used to monitor sensitive objects, the privacy of monitored objects' locations becomes an important concern. When a sensor reports a monitored object by sending a series of messages through the sensor network, the route these messages take in theory creates a trail leading back to their source. By eavesdropping on communications, an attacker may be able to move from node to node to follow this trail. Several approaches aimed at discouraging this kind of eavesdropping have been proposed, including mechanisms for constructing "phantom" routes and approaches that insert fake sources as background noise. A problem with existing approaches is that message latencies become larger and energy costs become higher as a result of introducing protections for the privacy of a source location. This paper proposes a new cyclic entrapment method (CEM) that protects source locations in sensor networks while adding a comparatively low cost in terms of additional message latency and energy.

All previously known algorithms for solving the multicommodity flow problem with capacities are based on linear programming.The best of these algorithms [14] uses a fast matrix multiplication algorithm and takes O(k25n2m5 log(nDU)) time to find an approximate solution, where k is the number of commodities, n and m denote the number of nodes and edges in the network, D is the largest demand, and U is the largest edge capacity. Substantially more time is needed to find an exact solution. As a consequence, even multicommodit y flow problems with jnst a few commodities are believed to be much harder than single-commodity maximum-flow or minimum-cost flow problems.In thk paper, we describe the first polynomial-time combinatorial algorithms for approximately solving the multicommodity flow problem. The running time of our randomized algorithm is (up to ,log factors) the same as the time needed to solve k single-commodity flow problems, thus giving the surprising result that approximately computing a k-commodity maximum-flow is not much harder than computing about k single-commodity maximum-flows in isolation.In fact, we prove that a (simple) k-commodity flow problem can be approximately solved by approximately solving O(k log2 n) single-commodity minimum-cost flow problems. Our k-commodity algorithm runs in O(knm log4 n) time with high probability.We also describe a deterministic algorithm that uses an O(k)-factor more time. Given any multicommodit y flow problem as input, both rdgorithms are guaranteed to provide a feasible solution to a modified fIOW problem in which all capacities are increased by a (1 + c)-factor, or to provide a proof that there is no feasible solution to the original problem. We also describe faster approximation algorithms for multicommodity flow problems with a special structure, such as those that arise in the 'sparsest cutn problems studied in [8, 10, 9], and the uniform concurrent flow problems studied in [13, 9] if k < fi.